Menu
Compact the current conversation into a handoff for another agent to pick up.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Deep Supabase diagnostics: pg_stat_statements for slow queries, lock debugging with pg_locks, connection leak detection, RLS policy conflicts, Edge Function cold starts, and Realtime connection drop analysis. Use when standard troubleshooting fails, investigating performance regressions, debugging race conditions, or building evidence for Supabase support escalation. Trigger: "supabase deep debug", "supabase slow query", "supabase lock contention", "supabase connection leak", "supabase RLS conflict", "supabase cold start".
Diagnose and fix Supabase errors across PostgREST, PostgreSQL, Auth, Storage, and Realtime. Use when encountering error codes like PGRST301, 42501, 23505, or auth failures. Use when debugging failed queries, RLS policy violations, or HTTP 4xx/5xx responses. Trigger with "supabase error", "fix supabase", "PGRST", "supabase 403", "RLS not working", "supabase auth error", "unique constraint", "foreign key violation".
Avoid and fix the most common Supabase mistakes: exposing service_role key in client bundles, forgetting to enable RLS, not using connection pooling in serverless, .single() throwing on empty results, missing .select() after insert/update, not destructuring { data, error }, creating multiple client instances, and not using generated types. Use when reviewing Supabase code, onboarding developers, auditing an existing project, or debugging unexpected behavior. Trigger with phrases like "supabase mistakes", "supabase anti-patterns", "supabase pitfalls", "supabase code review", "supabase gotchas", "supabase debugging", "what not to do supabase", "supabase common errors".
Scale Supabase projects for production load: read replicas, connection pooling tuning via Supavisor, compute size upgrades, CDN caching for Storage, Edge Function regional deployment, and database table partitioning. Use when preparing for traffic spikes, optimizing connection limits, setting up read replicas for analytics queries, or partitioning large tables. Trigger with phrases like "supabase scale", "supabase read replica", "supabase connection pooling", "supabase compute upgrade", "supabase CDN storage", "supabase edge function regions", "supabase partitioning", "supavisor", "supabase pool mode".
Optimize Supabase query performance with indexes, EXPLAIN ANALYZE, connection pooling, column selection, pagination, RPC functions, materialized views, and diagnostics. Use when queries are slow, connections are exhausted, response payloads are bloated, or when preparing a Supabase project for production-scale traffic. Trigger with phrases like "supabase performance", "supabase slow queries", "optimize supabase", "supabase index", "supabase connection pool", "supabase pagination", "supabase explain analyze".
Build resilient Supabase integrations: circuit breakers wrapping createClient calls, offline queue with IndexedDB, graceful degradation with cached fallbacks, health check endpoints, retry with exponential backoff and jitter, and dual-write patterns for critical data. Use when building fault-tolerant apps, handling Supabase outages gracefully, implementing offline-first patterns, or adding retry logic to SDK calls. Trigger with phrases like "supabase circuit breaker", "supabase offline", "supabase retry", "supabase health check", "supabase fallback", "supabase resilience", "supabase dual write", "supabase outage".
| name | handoff |
| description | Compact the current conversation into a handoff for another agent to pick up. |
| argument-hint | What will the next session be used for? |
Write a handoff summarising the current conversation so a fresh agent can continue the work, conversation will be compacted.
Include a "suggested skills" section in the document, which suggests skills that the agent should invoke, and include the current scope of work and what's next, and what principles or workflows are established with the user in current conversation that should carry forward.
Do not duplicate content already captured in other artifacts (CLAUDE.md, AGENTS.md), only reference and guide the agent on where to look, do not bloat the agent's context window.
Redact any sensitive information, such as API keys, passwords, or personally identifiable information.
If the user passed arguments, treat them as a description of what the next session will focus on and tailor accordingly.
